I-Like-To-Bike

Those alleged "reduced costs" apparently do not include the costs of owning and operating Uber's fleet of cars that is currently paid by its drivers, and with no drivers, the fixed costs (purchase/financing, insurance, registration, etc.) and variable costs (fuel, maintenance, tolls, parking/storage, etc.) of the fleet will be instead incurred by Uber and presumably those costs will need to be recovered from the passengers of the fleet.

It is Uber's investors' dream that they will eke out a profitable enterprise with driverless cars, since they currently are loosing hundreds of millions of dollars every quarter under their present scheme of operation.

I suspect that the only serious "profit" that will ever be made by Uber is by insiders/founders and early investors who can cash out after a glitzy IPO offering and exit the scene before Uber collapses from its money losing operations with or without driver less cars.

genec

Well it appears we both somewhat agree that any cost savings, if it might exist, is not likely to be passed on as lower fares to passengers.

I-Like-To-Bike

Dunno, Uber has kept their fares artificially low for years, subsidizing the fares with stupendous losses, and penalizing their drivers' ability to make a living (hence the constant turnover) all in the quest of gaining market share. They might keep fares low,and continue operating at a loss right until they bankrupt the company, or at least until the anticipated IPO payday for the Uber insiders.

It remains to be seen whether Uber will keep market share and its passengers if and when it ever charges a fare rate that will ensure profitability and a return on investment in, and operation of, a fleet of rental vehicles.

Ninety5rpm

Okay, so Waymo partners with Lyft and starts earning huge profits by not lowering its fares. If fares are not lowered, how does this even affect Uber, much less take it out of existence?

Ninety5rpm

This is how it looks to me: Waymo is going to be ready with Level 4 AV tech way before anyone else. It's a big investor in Lyft and the partnership is made - so Uber is screwed. Waymo's parent Alphabet also has billions in cash ready to be invested in myriads of cars.

Fares WILL be lowered and, one market at a time, Lyft using Waymo will drive Uber and all taxis out. Human-driven taxi/hailed ride business will go from viable to zero in a matter of weeks, if not days, once Lyft AV service hits an area.

Unless Apple or GM/Cruise pull off something amazing and partner with Uber, like soon, it's a done deal. We'll know for sure by 2020, if not sooner.

I-Like-To-Bike

Who is going to manufacture and sell the vehicles to Waymo/Lyft and at what price that allows Waymo/Lyft to earn "huge profits"?

What makes you think people cannot resist paying high fares for the pleasure of being chauffeured about daily with strangers in a driverless shared rental car?

squirtdad 

at least on startup is pushing the idea that a remote operator can take over in places where the AV system is at a loss

https://venturebeat.com/2017/12/06/p...s-in-a-pickle/

genec

Amazon did the same thing for years... and seems to be doing quite well now... has actually redefined the market place. Uber may do the same thing... or Weymo, if Uber can’t pull it off. Uber is NOT the only actor in this field... keep that in mind.

Ninety5rpm

There is no limit to dumb ideas.

Ninety5rpm

What do you mean who is "going to"? They've already bought hundreds of Chrysler/Fiat minivans, and are buying thousands more.

https://www.theverge.com/2018/1/30/1...n-self-driving

And it's not an exclusive deal. They can buy myriads more from anyone.

Rental car companies today are viable renting cars at $40/day. It doesn't take very many rides at 30 cents a mile to bring in comparable revenue. Anything on top of that is gravy.

I-Like-To-Bike

Uber might, Waymo might, some other organization or combination of them might be successful. Anything is possible especially predictions that are vague and have open ended timelines for producing a profitable operation.

However huge amounts of money spent, or trees cut down/electrons excited in producing hyped up press releases are no guarantee of success of a profitable enterprise.

I-Like-To-Bike

And none of them are self driving cars capable of hauling around paying passengers anywhere but on a test track or handful of isolated routes in a few locations?

What is the current price tag for these pseudo AV test cars? How many coins have "they" been paid by customers who seek this level of service?

Has a single person, anywhere, reduced their use of other means of travel (car, train, bus) due to a future purchase of hundreds or even thousands of Chrysler/Fiat minivans for testing by Waymo?

Where exactly have you seen reports of the humongous demand by the public for driver less cars or for a desire to replace the convenience of personally owned vehicles in order to share rides with strangers for every trip they take?

noisebeam

How does an AV make a left turn across a stop and go line of dual lane traffic in rush hour? This always requires high assertiveness (bordering on aggressive) and communication and cooperation from other drivers to open a gap. There is no waiting for a safe gap unless one waits an hour for rush hour to be over.

KD5NRH

And my mom just recently retired from a long and busy career finding the mistakes those computers are still making. You think companies just don't bother fixing the code because it's cheaper and easier to have a team of six-figure accountants finding $20-150 million in errors every year than to debug it properly? Plus, most of that stuff remains reversible for quite a while, whether by simply fixing the numbers in the computer or reshuffling money through the banking system, whereas when a car makes a bad decision, it happens within seconds and you can't just unwreck it and unkill the victims.

I-Like-To-Bike

Possibilities for current test models:
1. Human intervention by test operator either in car or elsewhere;

2. Selection of test routes to avoid all such intersections during rush or other busy hours, sort of the same way it handles snow and ice, make pretend it doesn't exist;

3. Handle like any other situation that it cannot cope with, come to a stop and simply and "safely" wait until the coast is clear no matter how long it takes and consider it proof of the safety of its non driver/software operation that doesn't suffer from driver aggression and impatience.

Ninety5rpm

If they don't believe they have the tech and ability and market demand to pull this off, why do you think they're buying all these minivans? Do you think you're smarter than Waymo engineers and Alphabet/Google/Waymo execs?

As far as I know they are capable of driving everywhere they've been taken so far, which includes dozens of US cities.

The price of the vans doesn't matter much, whether it's $40k or $100k - the revenue from these things will dwarf the purchase price either way.

The relative values of hailed rides vs driving your own car varies based on several factors, but most notably:

1. Amount of fare for the ride
2. Cost and ease of parking own car

The higher the cost and lower the ease of parking your own car, the higher the tolerance for hailed ride prices. Conversely, the lower the fares, the higher the willingness to use them instead of one's own car.

This is why hailed rides are more popular where parking is more expensive and more difficult. They're ubiquitous in places like San Francisco.

But when the fares drop, the balance changes, and demand for hailed rides increases. That's Econ 101.

Maelochs

yeah ... ummm ... sorry you Totally missed the point.

But I am pretty sure the idea that one cannot unkill people will prove useful.

Maelochs

the evidence of what some folks think is ... well they post their thoughts here.

It has been pointed out many times that the tech and most of the software issues have already been solved in different applications ... they fly in auto piloted airplanes while using smartphones and tablets to post that it is to have cars use the same tech.

But ... until AI cars either mostly take over or mostly fail ... not much point int he two sides debating. it is a "religious" issue now, where faith and belief is all that matters .... and the harder you push a believer, the harder they cling to their ideas.

On both sides. The people to whom AI cars seem like an obvious and not particularly difficult next step are as adamant as the people who are sure it will never work. That is why all this is pointless.

Maelochs

As for AV taxis ... Taxis.

Taxis work ... they generate enough income to pay for the car and the driver. An AI car only has to pay for the car. Since the operating cost will be lower, the companies will charge less ... not to be nice folks, but because they will undercut each other to get traffic.

It is capitalism 101 ... but obviously a lot of the urban planner/electrical engineer/software expert/vehicle designer/anthropologist/psychologists around here are also business majors.

I-Like-To-Bike

From this non-response I will assume that you have not seen any reports of a large demand from the public for AV cars or taxis.

The price of the vehicles doesn't matter? Is that also Econ 101?

Ninety5rpm

They will charge as little as possible not only to undercut others, but to effectively compete with people choosing to drive their own cars.

Even a monopoly AV hailing service would benefit from higher volume vs people driving their own cars. The more people using their service, the more flexible and convenient they can make their pooling services, and pooling services are more lucrative. I mean, carry one person for $10 or 6 people for $3 each... simple math...

[MENTION=20232]I-Like-To-Bike[/MENTION] and others... I've only started watching this preso from Waymo, but it's excellent.

https://www.youtube.com/watch?v=Uj-r...ature=youtu.be

If nothing else, watch this part about bike salmon detection in the dark:

https://www.youtube.com/watch?v=Uj-r...tu.be&t=24m58s

This video was linked in this excellent hot-off-the-press paper on AVs and detecting peds and cyclists by Dr. Stephen Goodridge (@sggoodri;).

Autonomous Driving and Collision Avoidance Technology

Ninety5rpm

Right. No reports for measuring demand of not-yet-available products or services. There were also no reports of a large demand from the public for cars, PCs, cell phones, online shopping or smart phones just before they became available. So what? What's your point?

I said it doesn't matter much. [MENTION=423651]Maelochs[/MENTION] explained why:

Capisce?

Ninety5rpm

The ability of AV tech to detect cyclists varies considerably based on how it is done. A thorough examination of the state of the art:

FEBRUARY 7, 2018

Autonomous Driving and Collision Avoidance Technology
Implications for Bicyclists and Pedestrians

Autonomous Driving and Collision Avoidance Technology

I-Like-To-Bike

Agreed, this is an informative wrap up on the current state of autonomous vehicles.
It may include the answer to Noisebeam's question about how they are programed to handle difficult left turns in rush hour:Dr Goodrich also cited the issue of the lack of public transparency of the testing process:

squirtdad 

why is it dumb? No matter how good AV gets there will be situations it cannot handle or handle well, this would be an option for those situations.

Whether is is marketable or successful or has a critical mass for success are valid questions.

There are so many companies in or jumping into this technology space it is mind bogglling. (good for my company because our chips are good fit for this problem)

this will probably sort out in the end to 2 or 3 technologies (think mac vs pc or adroid vs iOS)